Galaxy also gets into the act with marginal seat polling, predicting a net 18 seats to the Coalition. Mumble has a summary on his blog.

It shows Labor losing 6 seats in New South Wales, 10 in Queensland,  2 in West Australia and 1 in each of Tasmania and the Northern Territory, while picking-up 2 in Victoria. The swings it predicts for each state are broadly what you would expect, given other evidence we have, such as this recent compilation from Newspoll.

The swing is calculated, as with Newspoll, by projecting the swing in the seats sampled to the whole state. I don't think that is very valid methodology, for reasons explained in my previous post.

There may be a little smaller swing for marginals in New South Wales, Victoria, South Australia and West Australia, and a larger one in Queensland than for the whole state, or there might not be. Seems to me we're no more certain with this analysis of what is going on in the marginals than if we just applied state wide swing figures and left it at that.

Share this article on your favourite social bookmarking sites: